Hi, monitoring 7000+ hosts with a bunch of service checks on top is easily possible with Icinga. I’ve seen far more clients managed by single Icinga instances. Of course, you’ll need Icinga satellites. A single master node won’t be able to handle that load. And take care how often you schedule each check. Not every check needs to run every 5 minutes. Just do the math how many services you have per minute in average, times the number of hosts... and you’ll get an estimate of the network workload your satellites will need to handle.
Good infrastructure planning is required for this kind of size, naturally. Master and satellites should have backup nodes in place. Dedicated hardware also for satellites is recommended, or nodes that don’t have a high workload usually. (Icinga satellites combined with xcat satellites seem to work fine, for example). Make sure you do monitor your monitoring system. Mit freundlichen Grüßen / Best regards Jonathan Hermann CI Operations - Server Services IBM AIX (CI/OSB5) Robert Bosch GmbH | Postfach 30 02 20 | 70442 Stuttgart | GERMANY | www.bosch.com<http://www.bosch.com> Tel. +49 711 811-57169 | Mobil +49 152 07563620 | jonathan.herma...@de.bosch.com<mailto:jonathan.herma...@de.bosch.com> Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000; Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Dr. Volkmar Denner, Prof. Dr. Stefan Asenkerschbaumer, Dr. Michael Bolle, Dr. Rolf Bulander, Dr. Stefan Hartung, Dr. Markus Heyn, Dr. Dirk Hoheisel, Christoph Kübel, Uwe Raschke, Peter Tyroller From: icinga-users <icinga-users-boun...@lists.icinga.org> On Behalf Of Szabó Gergo Sent: Donnerstag, 30. August 2018 18:03 To: Icinga User's Corner <icinga-users@lists.icinga.org> Subject: [icinga-users] Generate Icinga notification based on Grafana/Prometheus data Hi, We have a lot of hosts (~7000) monitored by Grafana/Prometheus but there is an intention to connect these systems with our Icinga (or Icinga2), actually generate notifications from collected/displayed data. Icinga instances (mainly scheduler) already reach their limit at 3000-3500 check count, so currently it is impossible to monitor above mentioned 7000 hosts with them. However, it would be fine to send notifications by SMS or e-mail and show/retrieve network related problems in one alert history. Is it possible somehow? I have searched for it with any success. Maybe we will need another external application to do it but there is a lot of them and I don't know which one would be the close best option (if exists). Thank you in advance. Regards, Gergo
_______________________________________________ icinga-users mailing list icinga-users@lists.icinga.org https://lists.icinga.org/mailman/listinfo/icinga-users
